LIVE PULSE
4.9 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.6 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.4 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src2.2 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.7 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.4 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.3 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.3 CoMem Paper Proposes Shared and Individual Memory Design for LLM Multi-Agent Systems — 1 src1.3 Paper proposes evolving context parameterization for large language models — 1 src1.3 Fine-Tuning Vision-Language Models with Listener Gaze for Referring Expressions — 1 src4.9 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.6 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.4 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src2.2 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.7 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.4 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.3 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.3 CoMem Paper Proposes Shared and Individual Memory Design for LLM Multi-Agent Systems — 1 src1.3 Paper proposes evolving context parameterization for large language models — 1 src1.3 Fine-Tuning Vision-Language Models with Listener Gaze for Referring Expressions — 1 src
HEATPULSEAI MAGAZINES
FLIP · FOLLOW · SAVE
Log inSign up

preventative steering

topic1 events
papersSEP 10 04:00 UTC

Study Examines How Preventative Steering Defenses Hold Up During Adversarial Fine-Tuning

New arXiv research explores how language models resist harmful behavior shifts caused by malicious fine-tuning. The work evaluates preventative steering, a training-time method that injects undesirable persona vectors during fine-tuning and removes them at inference, and analyzes how the defense's effectiveness changes over the course of training. The findings suggest these safeguards require active adjustment across training phases rather than a fixed configuration.

safety/llm-security/fine-tuning/research